The design of a component to elicit broadly neutralizing antibodies is thought to be crucial for a successful HIV vaccine but is proving very difficult. In particular, immunogens designed from study of broadly neutralizing human monoclonal antibodies have generally failed to elicit antibodies with the appropriate specificities. A number of hypotheses have been advanced for these failures including restricted access to broadly neutralizing epitopes in recessed sites on the virus surface, proximity of the epitopes to viral membranes and epitope mimicry of self-antigens. In this new R01 proposal, we propose to generate transgenic mice carrying genes encoding two of the broadly neutralizing HIV antibodies 4E10 and b12 to investigate these hypotheses and to provide tools to study and rank potential HIV vaccine candidates. [unreadable] [unreadable] In the first instance, we propose to use mice carrying HIV antibody genes targeted to the physiological Ig loci to test if 4E10 and b12 have biologically relevant self-reactivity. If self-reactive, the B cells generated in 4E10 and b12 transgenic mice will be regulated by the immune tolerance mechanisms of receptor editing, anergy induction, or deletion and these mechanisms will be dissected. This may allow the design of immunogens in the future that do not evoke self-responses. If non-tolerant, transgenic mice carrying the 4E10 or b12 specificities should carry a greatly increased precursor frequency of B cells that are broadly neutralizing for HIV. We will then use an array of immunogens to explore the origins of any restrictions on eliciting 4E10 and b12 antibody specificities, to evaluate the most promising immunogens and ultimately to guide the design of modified and new immunogens to elicit broadly neutralizing antibodies. [unreadable] [unreadable] Relevance: To be successful, it is likely that an HIV vaccine must stimulate neutralizing antibodies. This proposal provides a mouse model to understand some of the problems in stimulating such antibodies by various vaccine candidates and to thereby suggest how better candidates can be designed. [unreadable] [unreadable] [unreadable]